Increasing the number of health plans available to the consumer has been embraced in both public programs and the private sector. Although some choice can benefit consumers, the literature on bounded rationality calls into question whether increasing consumer choice is always beneficial. Even for simple, everyday goods, cognitive constraints often overwhelm consumers facing a handful of choices. As the number of options increase, consumers tend to a) defer making choices they would otherwise find beneficial, and b) commit errors in the choices they do make. If the results are replicated for health plans, policies meant to increase consumer choice might actually: a) discourage choice, b) decrease consumer disenrollment from poorly performing plans, and c) increase errors in judgments when selecting among plans. To better understand the impact of choice sets on the selection of health plans, the proposed project would blend econometric and experimental methods. The experimental protocol (N=550) would help to identify if and when a larger set of health plan choices induces choice deferral or other associated choice errors. It would also identify the reasoning behind these patterns, suggesting policy interventions to enhance choice. The econometric work would then test for the existence and magnitude of similar choice pathologies in the Medicare population, drawing data from the Medicare Current Beneficiary Survey from the years 1995-2004. This work would examine how the number of Medicare Managed Care Plan options affects the propensity of a) those recently eligible for Medicare (N=6,500) to enroll in a Medicare Managed Care Plan, and b) those dissatisfied with their current Medicare Plan (N=4,375) (or Medicare Fee-For Service (N=4,250)) to switch plans (or enroll in a plan from FFS). Enrollment decisions are important to examine, since larger choice sets are intended to increase enrollment. Disenrollment is important to study because it serves as a) a means of protecting consumers who have had problematic experiences, b) a way to provide feedback to plans, and c) a quality indicator to other consumers (reported on the Medicare Plan Finder website). [unreadable] [unreadable] [unreadable] [unreadable]